March Madness: Table's Set for the Sweet 16

With its share of buzzer-beaters and upsets, the NCAA basketball tournament is wending its way to another memorable year. The decision to continue the tournament during the war is providing a needed diversion for both sports fans in this country and the troops fighting in Iraq.

In the penultimate weekend of the tournament we have an interesting array of games, although the point spreads for the matchups are larger than would ordinarily be expected in the third round. Part of my goal during this round is to avoid being too impressed or disappointed with performances in the first two rounds.

Because we have a limited number of games to bet on, the "public" money will be coming in on the teams looking impressive in the first two rounds. As a bettor, my goal is to zag when everyone is zigging and zig when everyone is zagging. Below are the matchups for this week's games. I've incorporated my power ratings into the charts this week, and the power ratings can be used to project the lines for Saturday and Sunday's Elite 8 games.

Teams	Official Line	Power Ratings	Lieberman Line	Difference
Thursday's Games
Wisconsin		10		2
Kentucky	12	0	10
Marquette		8		2
Pittsburgh	5	5	3
Notre Dame		10
Arizona	8	1	9	1
Duke		6
Kansas	3	3	3
Friday's Games
Auburn		14
Syracuse	5	8	6	1
Butler		14
Oklahoma	7 1/2	6	8
Connecticut		8
Texas	3	5	3
Michigan State		10
Maryland	3	6	4	1

I always find it useful to see if there are any trends at work in a particular round of the tournament to see if the trend will provide a discernible edge. For the third and fourth rounds, here are the statistics for the last six years.

Favorites in Third and Fourth Round Against the Spread
Year	Third Round	Fourth Round
2002	3-5	2-1-1
2001	6-2	2-2
2000	3-5	2-2
1999	2-5	1-3
1998	2-5-1	1-3
1997	4-4	4-0
1996	5-3	2-2
Totals	20-26-1	12-11-1

With the exception of 2001, underdogs have fared well in the third round, and I think they'll fare well this year, too. I say that because I believe the lines have been inflated on teams that have gone through the first two rounds without any hiccups, such as Kentucky and Pittsburgh. But in sports betting, there will generally be a regression to the mean, because the point spread on a team playing well becomes more onerous as a team's success grows.